Latest / Speaking of Supply Chain / Unlocking Efficiency and Ergonomics, with Sascha Kaczmarek of Motion Miners

Transcript

Unlocking Efficiency and Ergonomics, with Sascha Kaczmarek of Motion Miners · Speaking of Supply Chain

0:06
You're listening to Speaking of Supply Chain, a meboch podcast. This is a
0:10
show for logistics professionals looking to learn more about the latest innovations in supply chain.
0:16
Each episode will feature a conversation on topics such as mitigating supply chain disruption
0:20
and reducing risk, current automation trends, sustainability initiatives, and more. Let's
0:26
dive right in. Hello, and welcome to Speaking of Supply Chain, where
0:31
we explore trends, current events, and innovations impacting the logistics and supply chain
0:36
industries. I'm your host, Ellen Wood. I am joined today by Sasha
0:41
Kathmeric, one of the founders of Motion Miners, and we are going to
0:46
be talking about process optimization and wearable technology that supports these initiatives. Welcome Sasha,
0:53
Hi Ella, and nice to meet you. Really looking forwarding the podcasts.
0:57
Fan at Stick, I think we have a great conversation today because wearables
1:03
and like the telemetry that we can get from these is so interesting. I
1:07
can't wait to hear more about it. So onto our conversation today, which
1:11
is about wearable technology and the innovations that have come about that use this technology
1:19
not just for personal information I know lots of people are familiar with the Apple
1:25
Watch, the fitbit, there are lots of trackers. I wear one,
1:27
and some of the things that we're able to capture about our own health and
1:33
wellness as we go about our day. I know I had a very busy
1:37
weekend this past weekend, and on Saturday, I was doing yardwork and I
1:41
ended up walking over sixteen thousand steps that day, which is a really big
1:45
day for me, and I certainly felt it the next day. So what
1:51
types of how our businesses trying to use this and use this in a commercial
1:57
application? Going on? Yeah, yeah, you mentioned it totally right.
2:04
So in the B two C context, the focus is definitely on your health
2:07
data, improving your fitness, and maybe also to have an easier life by
2:13
connecting it to your mobile devices for easier handling of your messages and calls and
2:17
so on. So this is usually the focus if you're wearing the Apple Watches,
2:23
the other smart watches and fitness trackers and so on. But using variabilts
2:28
in a B to B context is a little bit different because in this context,
2:31
it is not interesting to understand what a certain employee, for example,
2:38
is doing the whole day, but it's more for a use to understand how
2:43
processes works, for example in production or logistics, because in many different production
2:49
logistics systems you have a lot of manual processes and you do not know what
2:53
is really happening on the shop floor. And if your processes are for example,
2:59
une efficient or unergonomic, then you will lose a lot of money.
3:04
Off course, and with the help of industrial wearables which look a little bit
3:07
different than yeah, commercial wearabils, you are able to analyze this menual shop
3:13
flow processes in an automated way and by this you can find potentials to improve
3:17
on the one side, the efficiency or the ergonomics. So if you compare
3:23
to be to see the focus is more on the process not on the person,
3:27
especially because of the data security laws in different countries, because you're not
3:30
allowed to understand what a certain person is doing in your processes. And from
3:36
a technical point of view, the weariabils are quite similar. From the technical
3:42
point what is inside So you have this guroscope, you have this extrometer,
3:46
a sensor insight and by how the sensor is moving you understand which process took
3:52
place currently. But the Yeah, the disadvantage of commercial wearables currently is that
4:00
they try to save better lifetime because you want to use them over a week
4:03
if possible, and therefore they do not store the data often enough on the
4:09
devices itself. But we need a lot of data to do this process analysis
4:14
in an industrial fields. So they are the difference and this is the reason
4:16
why we cannot just use a smart watch and apple water or fitness tricker to
4:20
do this industrial use cases, which we are currently doing. Okay, so
4:26
focusing on some of these these industrial wearables, you said they're not the same,
4:30
they're they're very different. So a watch wearable is or or a watch
4:35
type wearable that would go on the rest is one thing because of the motion
4:40
that's that's needed to track. What other components are there to these wearables.
4:46
Yeah, for example, what we need we need at least three wearables,
4:49
so one on each wrist and one the belt because one is currently not enough
4:55
to do this activity recognition. We have different machine learning igorithm we're able to
5:00
do these activity recognition algorithms and to have a good valid database in the end,
5:05
we need at least three sensors. There are also some other companies which
5:10
are using some sensor suits with a lot of more sensors to get much of
5:15
more deeper insights in your into your processes. But unfortunately, if you have
5:18
to wear a sensor suit with at thirty degrees celsius over two or three weeks,
5:25
it's not that cool for your employees. So this is the reason why
5:28
we're focusing on three sensors. And what we use in additional are are so
5:32
called bluetooth beacons, and these bluetooth beacons are used to create an indo or
5:38
indolocalization system. This is important to understand where the process took place because we
5:46
cannot use the GPS signal in a warehouse or in a production system because we
5:50
are insight So this is the issue that we need extra or extra hardware to
5:57
get this indo localization system. So kinds of data do you capture? You
6:01
mentioned having you know, process data, but in terms of what these wearables
6:08
are capturing about, like space and movement and speed, what kinds of things
6:14
are you looking for and what can you do with that information? Yeah?
6:18
Yeah, So the raw data itself is just information about exomation and how the
6:25
movement looks like by the ugerroscope, and then we have an algorithm who is
6:30
able to understand if this data looks like walking, standing, driving, handling
6:35
things, bending, pushing, pulling a card or carrying a carton and so
6:39
on. So we trained the algorithm to understand how sensor data looks like if
6:45
you move in a certain way. And this is the base activity data which
6:48
we are focusing on. And then in addition to that, we have this
6:53
localization data based of the bluetooths beacon signal which is gased by these sensors for
6:59
the information about the where the process took place. And what we then do
7:02
is we can combine the two worlds together. So we use the activity data
7:09
and we use the location data and create KPIs out of it. For example,
7:14
we have the KPIs like waiting time, the number of congestions or the
7:18
congestion times, the walking or driving times, the walking distance, the number
7:24
of bending movements, and so on and so on. So we try to
7:28
get this KPIs which you usually have to gather if you use manual process studies
7:32
like stopping these times with a time stop up. This is what we try
7:36
to get out automatically out of the sensor data. And then based on the
7:41
sensor data the process engineer or the industry engineer is able to understand the process
7:46
much more better, to see where bottlenecks in my systems, where can I
7:48
improve the ergonomics, for example for my employees, and then to think about
7:53
optimization measures which can then be implemented into the process. So do you certain
8:00
industries have more interest in this type of data or technology or is it more
8:07
just function based where it's you know, a more manual operational, more manual
8:11
warehouse or production something of that nature. Does it does it cross industries or
8:15
are there certain industries that this is this is very popular right now. Yeah,
8:20
it is cross industries definitely. But we always say industries are good for
8:26
the technology which have a lot of sites, with a lot of manual processes,
8:31
with a lot of employees which are still doing the hard work every day,
8:35
and there we can help them with the with the technology. So for
8:39
example, our main companies are from the field of logistics service provide us like
8:46
a DHL or dB Shanker for example, or from the retail companies in Germany
8:52
for example, like a Leader, or in the US hopefully in the future
8:56
Walmart. So this is comparable. But we have also customers from from the
9:01
field of manufacturing, automotive suppliers or the automotive company itself. So we always
9:07
say everywhere where manual processes take place, there we try to help with our
9:13
technology. Sometimes we have really exotic questions which which prop up from from customers.
9:20
We had some I would say POC's in hospitals where we analyze hospital processes
9:26
also quite interesting to see. So I think there are no limitations. What
9:31
is important in the end is that we can deliver can deliver yeah, something
9:35
which what will help or which will help the customer to improve their processes in
9:41
the end. So when it comes to some of these these processes, is
9:46
it only about ergonomics or is it about actually improving and optimizing maybe their their
9:54
workflows or you know what is what is the impetus that gets companies interested in
10:00
this type of data gathering, because that's what this is. This is this
10:05
is gathering data to try to solve something. What are they usually trying to
10:09
solve? Yeah, most of the time it's both. But the most companies
10:13
of course would like to see how they can improve their processes regarding the efficiency.
10:18
So how we can set up lean up processes, how we can create
10:22
more output with my team, which I currently have because by this they can
10:26
save money in the end or more money. This is often the most important
10:31
part for them. But yeah, but the good thing is that based on
10:35
the same data without doing an additional analysis, we can do it in the
10:39
same time, or we can do an economic analysis at the same time,
10:43
so we have a win win win situation. We have a win situation for
10:48
the company or for our client because they can save money. We have a
10:50
win situation for the employee itself because we can try to improve their ergonomics.
10:56
And on our side, we have also a win situation because we're a little
11:00
bit of money. So the win win win situation in the end. Sure,
11:05
So what type of ergonomic situations are usually able to be resolved? I
11:11
know we have an initiative going on in our office right now where we're trying
11:16
to make sure that the ergonomics are right, that everybody's chair is the right
11:20
height and the distance from their computer and their desk, and we have standing
11:24
options and sitting options and lots of different ways to make sure that we are
11:28
all you know, appropriately, I don't even know, just positioned to do
11:35
our work. Most efficiently. But also that's for a health and wellness you
11:41
know aspect. So what types of things are are we able to in an
11:45
industrial situation improve in terms of those ergonomics. Yeah, so this is of
11:50
course a little bit different if you compare to your office processes. What we
11:56
often see in for example, inbound processes that they have to handle a lot
12:01
of goods in a banded position because there are no solutions to lift things up,
12:05
for example, or if we are in a picking process, we often
12:09
analyze how often they pick something over the head and in a banded position,
12:13
and then we think about if we can restore the articles in the warehouse or
12:18
that the a runner. So there the goods which which are sold the most,
12:22
that they are in the economic zone so that they don't have to pick
12:26
in an your bendet position anymore, or on a position where they have to
12:31
go with their arms over the head. So this is what we are focusing
12:35
on. And also the distance which is traveled through the system, because in
12:39
a lot of projects you can imagine definitely in picking processes like at Amazon,
12:45
for example, they travel a distance about twenty kilometers a day, and this
12:50
is from an economic point of view, not that good. On the other
12:52
side, standing the whole day is also not good. And then you have
12:56
to think about how it's right. And then you have to think about if
13:00
you can use jop rotation, for example, and by this you can reduce
13:03
the strain for the employees and you have a much more better process for them
13:07
and hopefully they will be much more healthy in the future. Wonderful well.
13:11
And so along with that ergonomic portion of it, so making sure that your
13:18
employees are working safely and not straining themselves and taking the breaks or rotating the
13:26
shifts like you were mentioning, to make sure that they're not loading all of
13:31
that activity onto their day and taking rests and things like that. So what
13:37
does that mean in terms of those processes that those companies have, because that
13:41
would be a big change that they have to do. If they have to
13:45
say, okay, well, they can't pick something else up off the floor
13:48
that's too much of a strain, or they can't take a heavy item you
13:52
down from above their head that's unsafe, or this repetitive motion that they're doing
13:58
for eight hours a day is too much. It needs to be in a
14:01
shift of four hours and then they switch to a different activity. So how
14:05
complicated does it get to change those processes after the fact and reimagine how their
14:11
business is going to function once they realize that perhaps this is too hard on
14:18
the employee's physical bodies. Yeah. Yeah, what our clients definitely have to
14:24
do is they have to invest something after such an analysis. But they are
14:28
open to invest something because otherwise they wouldn't do an analysis with us together or
14:33
do it by theirselves with our technology. So often after we analyze data,
14:41
we have Yeah, we do small quick wins, so quite simple things which
14:45
can improve the process directly. For example, one use case while at a
14:50
logistics service providers there we saw in data that the employees had to walk a
14:54
really long distance to a thirtain central printer. I don't know how many kilometers
15:01
a if you scale it up over twenty employees per ship, and then we
15:07
offered a quite simple solution for them just by ten more small printers, put
15:11
them at their working desk, and then they don't have to travel anymore to
15:16
this central printer, so that we can reduce the strain by walking this distance,
15:20
and in fact we in the same time we improve the efficiency because it
15:24
says of course a lot of time, so it's then bows again, right,
15:28
But sometimes it's a little bit more complicated. Of course, when we
15:31
see data, when we see something like the picking algorithm is not that good
15:37
or the storage algorithm is not that good, then you have to improve something,
15:41
maybe in your w mess system. And this sometimes a little bit more
15:45
effort for the IT guys, and it's costs a little bit more of money
15:48
to implement it. Or if we see in data that it makes sense to
15:52
use exoskelets, for example, and you can imagine if you roll out exoskelets
15:56
for your one hundred employees in a warehouse, it costs also a lot of
16:00
money. And the payback regarding efficiencies often later, maybe after five or ten
16:07
years, because you see it cannot see it in data after one or two
16:11
years, because the health data is approvable after roughly five to ten years,
16:18
and then you can see if you have to payback or not. So you've
16:22
you've mentioned a couple instances, you know, short snippets about what customers have
16:26
been able to achieve or understand about their operations. Do you have just like
16:33
an overall success story that tells more about like the scale of improvements that companies
16:40
can do. So when they're coming they're trying to make improvements, they're trying
16:42
to find those efficiencies and reduce their costs, improve their output. You know,
16:48
what's what's your best customer success story? Yeah, so there's a quite
16:55
new story. I think the project was from this year where we did pros
17:00
analysis at a fashion e commerce company and there we analyze their picking process and
17:06
the process took place in a five level wrecking system, so quite a huge
17:10
system with about twenty five employee per level, so it was really a huge
17:15
warehouse. And the goal was definitely to increase the efficiency and on this in
17:22
the same time or at the same time, the economics was the employees.
17:25
And what we did there was we analyzed the process for two weeks because after
17:30
two weeks we had enough valid data. In this two weeks we get said
17:34
over one year of process data, which no one will do if you do
17:38
it manually, so no one is running behind an employee one year because there's
17:42
cost too much money. And for example, what we've found out in this
17:48
data was that we had a long that we have many long searching times for
17:52
articles, for example, because of the WMS system how it was set up.
17:57
We had long walking disc insus sometimes they visit an aisle twice, so
18:03
this was also not good implemented in their WOMS system. And what we also
18:08
saw in the data was that we had many unergonomic pick processes and in addition
18:15
to that many clarification cases. And this means that sometimes they have some issues,
18:21
for example that they couldn't find the article or that the article wasn't on
18:23
stock anymore. And in this case or in the baddest case, they had
18:27
to go down from the fifth level to the low ground level and then go
18:33
to the shift leader, ask the shift leader what they can do, and
18:37
then they had to go back to the fifth level. And this of course
18:40
costs a lot of time to clear the clarification case. And this is something
18:45
we saw in data later on, and based on this data and the results,
18:51
yeah, we have developed several optimization measures for examples, such measures like
18:56
using a program of hand scanner to reduce the pick times, or use a
19:00
narrow picking card so or a smaller card for the picking process to avoid congestions
19:07
in their processes because the aisles were really small or really tight. And for
19:12
example, what we also saw that it would definitely make sense for the customer
19:17
to think about a better routing strategy in their WMS system, because there we
19:22
had the biggest improvement, but of course also the biggest investment, because then
19:26
you need the SAP guys who are working on the algorithm again and so on.
19:32
And what we saw later after calculating the saving potentials was that we could
19:38
save for that customer about three hundred thousand euros so currently also three hundred thousand
19:45
dollars per year. And at the same time we improved the economics by different
19:51
measures by twenty percent, So this was quite a gooding project for our customer,
19:57
significant and that was just in one of their facilities. That's a one
20:02
facility. Yeah, and then what you can see if you have best practices
20:06
and roll it out to the other facilities, and then you safe again three
20:08
hundred cells in the other site. So quite simple, right. If if
20:14
you're gonna if you're gonna change up your WMS system and you know, reprogram
20:18
some things, you might as well make these these changes system wide and make
20:22
sure that all of your employees are benefiting from the information that was gleaned at
20:26
that one site. Well, and that brought up another question to me.
20:32
So, as you're going through you're collecting the data. Obviously Motion Miners is
20:37
the company that has the wearables that does this. But then when it comes
20:41
to those suggestions or those those potentials for changing those processes, do you also
20:48
walk with the client through that process of this is what you need to do
20:52
instead or is it just this is where there's a problem, it needs fixed.
20:56
Here a couple solutions. But how long does that engagement last with Motion
21:03
Miners in terms of coming up with all of these KPIs and these new processes
21:07
that are going to be much leaner and much more efficient. Yeah. So
21:12
currently we have two business cases. On the one side, we do consultant
21:17
consultancy for our customers, so we set up the system, analyze the data,
21:22
interpret the data, and we create these optimization measures for them. And
21:26
sometimes we also start there, i would say, rough process planning and how
21:32
to implement the new organizational measures or the technical measures. But what we often
21:37
do not do is something like a fine planning of the system, or we
21:41
do not help them by realize all things because we are not the logistic planners.
21:47
We are process optimization process optimization company, and if our customers or clients
21:52
would like to, yeah, implement all the things which we suggest to them.
21:57
Then we have partners for that one who will will help them with it.
22:00
And on the other side or most of the time bigger clients can also
22:06
or will also license or technology, so there's a yearly licensed model behind it,
22:11
so they can license our hearted software. Then we train them and then
22:15
they can analyze their processes totally by their selves, and we help them also
22:18
to understand the data and to understand how they have to interpret this data to
22:23
come to an optimization potential. And this is also possible. And in this
22:29
case we have a customer success team which just helped them a little bit,
22:33
but in fact we try to onboard them so that they can do everything by
22:36
theirselves. What is the reaction from these employees on you know, in these
22:42
companies, on these production or warehouse floors. Are they are they open to
22:48
the concept of being tracked and looking for improvements? Are they hesitant? Do
22:56
they do they push back on it? Is there much change management that needs
22:59
to happen and sort of getting buy in for this study? Yeah, yeah,
23:03
this is one of the most important part. When we start a project
23:07
with a customer that we will be or that we are as much as transparent
23:14
as possible so that we explain to the employees and maybe also to the unions.
23:18
In Germany, you have these I don't know if you understand the word
23:22
the bitriepsrat if the German word, and this is really important for them that
23:27
they understand what we are doing, why we are doing it, and what
23:30
we are not doing and not allowed to do because of data security laws,
23:36
because there are a few things that we are not allowed to do, for
23:37
example, analyzing times when they are doing a break or when they smoke a
23:42
cigarette or going to the toilet for example, and these are the concerns they
23:48
have and this is typically change management what we have to do, and we
23:52
are really well trained in how to handle these issues. And one of the
23:57
also one of the most important point is that we have this data security or
24:03
data legal opinion which also confirms that in the way how we do this motion
24:07
mining analysis, that we are totally compliant, especially here in Europe with the
24:12
GDPR laws, which are i would say the most of the hardest laws regarding
24:18
data anonymization and data security laws and so on. And this is what we
24:22
deliver a solution where you cannot see later on in the data from which employee
24:26
the data comes from. This is really important. You are not allowed to
24:30
see that Missus Miller, for example, created this waiting or walking time.
24:33
This is not allowed. What we are only allowed to do is to create
24:37
process data in the end. And this is something you really have to explain
24:41
to the employees. Yeah, so then they know that this information is being
24:45
collected based on function and that employee function, not that employee specifically correct and
24:52
correct. Interesting and are you currently only working in Europe? Are you global?
24:56
Do you work in North America, Africa, Asia? Where where is
25:02
motion miners be able to support clients right now? Yeah? I would say
25:07
we can help them everywhere on the world. But of course we started in
25:11
Germany and Europe. This was the easiest start for us. We had one
25:17
project last year in Mexico automotive supplier. We have one license customer in Malaysia,
25:22
for example, and we will start this year to focus more on the
25:29
North American market. So I myself, I will be in the US from
25:33
beginning mid of October to the end of November because there's a trade fair industrial
25:41
transformation is it called I think in Indiana real transformation. Yeah, this year
25:45
in Indianapolis this year, right right, So currently we are talking yea to
25:52
the city of Dormont. Would will support us to get over the Big Sea
25:56
and participate at this fair and then hopefully we can participate also in this German
26:00
accelerator. This is a program for startups which will start in the US,
26:06
and then I will be on site for four to six weeks. And if
26:08
some of our listeners now is interested to meet me, he can just give
26:12
me a call or write me an email and then I will do on side
26:15
appointment. This is also no problem, fantastic. Well, now our listeners
26:21
know where to get more information about motion miners, so obviously there's your website.
26:25
You will be in Indianapolis this fall, so we will we will have
26:29
to make plans to grab a coffee together because there's only a couple of blocks
26:32
from where I'm sitting right now. That will be awesome and that's that's so
26:37
exciting that you'll be coming to North America. I'm sure our listeners will be
26:40
interested and we can certainly direct them to where you will be. And that
26:45
brings us to the end of this episode. Thank you so much Sasha for
26:49
telling us more about this wearable technology and what it can accomplish for companies,
26:55
whether it is a three pl or a fashion house or just a warehouse or
27:00
production facility, that they can get some of these optimizations and improvements in their
27:07
organization, both from a process standpoint, but also ergonomically for their employees to
27:14
make sure that their employees are safe and healthy on the job. I hope
27:18
our listeners enjoy the discussion as much as I did. If any of you
27:22
out there have a suggestion for a topic or would like to be a guest
27:26
on our show, please let us know we'd love to hear from you.
27:29
You can contact me at podcast at mebak dot com anytime. As always,
27:33
thanks for listening to speaking of supply chain. As always, thank you for
27:37
listening and tune in next time. Mebach Consulting is one of the largest and
27:45
most globally recognized supply chain consulting, engineering and advisory firms. For nearly fifty
27:52
years, we've helped clients achieved supply chain excellence and sustainable competitive advantage across the
27:56
entire spectrum of the supply chain by delivering improvements and innovation strategically tactically and digitally.
28:04
To learn more, visit meebok dot com. You've been listening to Speaking
28:08
of Supply Chain, a Mebok podcast, keep connected with us by subscribing to
28:14
the show in your favorite podcast player. If you like which you've heard,
28:17
please rate the show that helps us to keep delivering the latest in supply chain
28:21
information. Thanks for listening. Until next time,